You are viewing a plain text version of this content. The canonical link for it is here.
Posted to cvs@httpd.apache.org by Brian Behlendorf <br...@hyperreal.com> on 1997/02/03 03:08:18 UTC

cvs commit: apache/src http_core.c

brian       97/02/02 18:08:18

  Modified:    src       http_core.c
  Log:
  Reviewed by:	Dean Gaudet <dg...@arctic.org>, Jim Jagielski <ji...@jaguNET.com>
  Submitted by:	Ed Korthof <ed...@organic.com>
  
  Fix core dump problems with Rlimit directives
  
  Revision  Changes    Path
  1.63      +1 -1      apache/src/http_core.c
  
  Index: http_core.c
  ===================================================================
  RCS file: /export/home/cvs/apache/src/http_core.c,v
  retrieving revision 1.62
  retrieving revision 1.63
  diff -C3 -r1.62 -r1.63
  *** http_core.c	1997/02/01 22:03:36	1.62
  --- http_core.c	1997/02/03 02:08:16	1.63
  ***************
  *** 1004,1010 ****
    	return;
    	}
    
  !     if ((str = getword_conf(cmd->pool, &arg)))
    	if (!strcasecmp(str, "max"))
    	    cur = limit->rlim_max;
    	else
  --- 1004,1010 ----
    	return;
    	}
    
  !     if (arg2 && (str = getword_conf(cmd->pool, &arg)))
    	if (!strcasecmp(str, "max"))
    	    cur = limit->rlim_max;
    	else
  
  
  

Re: cvs commit: apache/src http_core.c

Posted by Ed Korthof <ed...@organic.com>.
Something went wrong -- this isn't exactly the patch I sent (at least not
as recorded in my sent-mail).  I'm not sure exactly what went wrong, but
the 'arg2 &&' should be added about ten lines below where this occurs --
this introduces an error where providing only one argument produces an
error message (and nothing is set).  (The person who submited the bug
applied the patch by hand, as this patch would do, and got that error --
could this patch have come from him?  The line numbers are right, from
the source code file he showed me...)

I don't know how to get a current snapshot, so I'll get one early tomorrow
morning and write a patch against it.  Alternately, I've included my patch
at the end of this message, in case it's easier to roll back one and apply
the other.  There are small but significant differences.  The line numbers
are a bit screwy, though, I guess because my patch was against slightly
older source...

     -- Ed Korthof        |  Web Server Engineer --
     -- ed@organic.com    |  Organic Online, Inc --
     -- (415) 278-5676    |  Fax: (415) 284-6891 --

On Sun, 2 Feb 1997, Brian Behlendorf wrote:

> brian       97/02/02 18:08:18
> 
>   Modified:    src       http_core.c
>   Log:
>   Reviewed by:	Dean Gaudet <dg...@arctic.org>, Jim Jagielski <ji...@jaguNET.com>
>   Submitted by:	Ed Korthof <ed...@organic.com>
>   
>   Fix core dump problems with Rlimit directives
>   
>   Revision  Changes    Path
>   1.63      +1 -1      apache/src/http_core.c
>   
>   Index: http_core.c
>   ===================================================================
>   RCS file: /export/home/cvs/apache/src/http_core.c,v
>   retrieving revision 1.62
>   retrieving revision 1.63
>   diff -C3 -r1.62 -r1.63
>   *** http_core.c	1997/02/01 22:03:36	1.62
>   --- http_core.c	1997/02/03 02:08:16	1.63
>   ***************
>   *** 1004,1010 ****
>     	return;
>     	}
>     
>   !     if ((str = getword_conf(cmd->pool, &arg)))
>     	if (!strcasecmp(str, "max"))
>     	    cur = limit->rlim_max;
>     	else
>   --- 1004,1010 ----
>     	return;
>     	}
>     
>   !     if (arg2 && (str = getword_conf(cmd->pool, &arg)))
>     	if (!strcasecmp(str, "max"))
>     	    cur = limit->rlim_max;
>     	else
>   
>   
>   
> 

*** http_core.c.orig	Wed Jan 29 21:34:44 1997
--- http_core.c	Wed Jan 29 21:34:17 1997
***************
*** 1003,1009 ****
  	return;
      }
      
!     if ((str = getword_conf(cmd->pool, &arg2)))
  	max = atol(str);
  
      /* if we aren't running as root, cannot increase max */
--- 1003,1009 ----
  	return;
      }
      
!     if (arg2 && (str = getword_conf(cmd->pool, &arg2)))
  	max = atol(str);
  
      /* if we aren't running as root, cannot increase max */